The end of his book "The Dilbert Future" calls for "new ways of looking at existing things." And then claims that all of science is just "looking at things" which, he points out, is pretty unreliable, and that we need to improve our ability to perceive the universe in order to move forward in science. Of note is that he doesn't consider technologies which enhance our perception to be actually doing that (microscopes, telescopes, and other types of sensors are just more forms of "seeing" to him, and are lacking).
He then rather poorly explains double slit experiment and suggests that the arrow of time might just be a figment of your imagination, as well as the motion of objects in general, and gravity.
He skirts around the idea that looking at existing ideas in new ways can be important (who would deny that?) but then devolves into useless examples, claiming that instead of gravity "existing", everything in the universe could just be expanding at just the right rate to create the illusion of gravity. He questions causal relationships and posits something no better than leibniz's monads.
I've derailed a bit here, but man, that book irked me.